Re: st: predict with xtreg

[Maarten Buis <maartenlbuis@gmail.com>]

On Thu, Nov 8, 2012 at 3:53 PM, Carmen Blanco wrote:
> According to your help the problem is because few variables in my
> model has a missing value for the observations. And , in fact, I want
> to say that my panel data is unbalanced. Do you know please how to
> calculate the random error component (u) in my panel data with another
> way?

Why would you want to do so? Those observations with missing values
weren't used, so -predict- is doing the right thing to ignore them.

If you want to use the observations with missing values you'll have to
make sure that -xtreg- can use those observations, e.g. -help mi-. Be
warned: doing that right is a major project, so plan a couple of
months minimum for that and ask yourself if 52 observations is worth a
couple of months of your live...

-- Maarten

---------------------------------
Maarten L. Buis
WZB
Reichpietschufer 50
10785 Berlin
Germany

http://www.maartenbuis.nl
---------------------------------
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/